Efeitos da fragmentação sobre a comunidade de lagartos em áreas de Floresta Estacional Semidecidual Submontana no Sudoeste de Mato Grosso, Brasil

Abstract

I evaluated the effect of fragmentation on the lizard community in seasonal, semideciduous, sub-montane forests in southwest Mato Grosso, Brazil. My study sites included 18 forest fragments between 41 and 470 ha in size and three control areas of over 1,000 ha. The fragments were isolated approximately 20 years ago and exist within a matrix of agricultural pasture. I used pitfall traps to census the lizard community of each study site. Five parallel lines of traps, spaced 50 m apart, were placed from the forest edge into the interior of each study site. An additional line was placed from the forest edge into the pasture, using the same spacing. Each line of traps was 50 m long, containing five 23.6 liter barrels, 5 m apart, linked by an 80 cm-tall fence. Visual samplings were also accomplished. I sampled the study sites between December 2002 and August 2004. Nine study sites were sampled twice, once during the dry season and once during the wet season. The rest of the sites were sampled only once during the study period, in either the wet or the dry season. For each sample the traps were opened for 10 consecutive days, for a total of 9,000 barreldays of sampling. I collected 322 individuals representing 7 families and 15 species over the entire study period. In 8 of the 9 study sites sampled twice, there was no effect of season in 2003, on the species richness and abundance of lizards. However, the species Mabuya nigropunctata was more abundant during the wet season. During the wet season, I only captured adult individuals of the species Stenocercus caducus, while during the wet season, most individuals of this species were juveniles. I also evaluated the influence of the matrix and the effect of the forest edge on the species richness and abundance of the lizard community. None of the species in my sample exclusively used the matrix habitat; most species showed a more generalized use of habitat. For example, Tupinambis merianae, Ameiva ameiva, Stenocercus caducus, and Cercosaura eigenmanni were captured from the matrix all the way through to the interior forest. Mabuya nigropunctata, Gonatodes hasemani, Iphisa elegans, Gonatodes humeralis, and Anolis fuscoauratus were captured from the forest edge to the forest interior. However, some species were encountered more frequently in certain habitat types. Colobosaura modesta, Hoplocercus spinosus, Bachia dorbignyi, Micrablepharus maximiliani and Kentropyx calcarata were only captured in the forest interior. I found significant differences in the species richness and abundance of lizards between the matrix and the forest edge and between the matrix and the forest interior. Lizard abundance did not vary between the forest edge and forest interior, but species richness was significantly different between the forest edge and 200 m into the forest. Looking only at fragments between 41 and 100 ha in size, species richness was significantly different at distances of 50, 100, and 200 m from the forest edge. An HMDS ordination did not show a grouping of study sites using either qualitative or quantitative data, indicating that there was no pattern in species composition among study sites. I performed multiple regression analysis to examine whether physical variables (fragment size, index of fragment shape, and time since isolation) and forest structure variables (number of trees in a fragment, canopy cover, and leaf litter volume) influenced lizard species richness or abundance. I found no significant effect of physical or forest structure variables on the overall species richness and abundance of lizards. However, forest fragment size had a significant positive relationship with the abundance of two lizard species, Gonatodes hasemani and Cercosaura eigenmanni. When I performed multiple logistical regression analysis, I found significant relationships between the abundance of Bachia dorbignyi and the number of trees and between the abundance of Anolis fuscoauratus and both the number of trees and leaf litter volume.